Metal-Induced Energy Transfer (MIET) Imaging of Cell Surface Engineering with Multivalent DNA Nanobrushes

The spacing between cells has a significant impact on cell–cell interactions, which are critical to the fate and function of both individual cells and multicellular organisms. However, accurately measuring the distance between cell membranes and the variations between different membranes has proven to be a challenging task. In this study, we employ metal-induced energy transfer (MIET) imaging/spectroscopy to determine and track the intermembrane distance and variations with nanometer precision. We have developed a DNA-based molecular adhesive called the DNA nanobrush, which serves as a cellular adhesive for connecting the plasma membranes of different cells. By manipulating the number of base pairs within the DNA nanobrush, we can modify various aspects of membrane–membrane interactions such as adhesive directionality, distance, and forces. We demonstrate that such nanometer-level changes can be detected with MIET imaging/spectroscopy. Moreover, we successfully employed MIET to measure distance variations between a cellular plasma membrane and a model membrane. This experiment not only showcases the effectiveness of MIET as a powerful tool for accurately quantifying membrane–membrane interactions but also validates the potential of DNA nanobrushes as cellular adhesives. This innovative method holds significant implications for advancing the study of multicellular interactions.


Apparatus
The cell fluorescence images were captured using a confocal laser scanning microscope (Nikon A1R) equipped with 4×, 20×, 60× and 100× objective lenses.All MIET measurements were carried out with a homebuilt confocal microscope equipped with a multichannel picosecond event timer (HydraHarp 400, PicoQuant GmbH) allowing for fluorescence lifetime imaging.Atomic force microscopy (AFM) characterization was carried out using a Bruker Dimension Icon (USA).Flow cytometry results were recorded on a FACS Calibur (BD, USA).Scanning electron microscope (SEM) images were measured on JEOL JSM-7500F (Japan).Cryo-TEM data were acquired using a spherical aberration (Cs) corrected FEI Titan Krios (Thermo Fisher Scientific) transmission electron microscope.
Table S1.Sequences of the oligonucleotides used in this work.The green and yellow parts represent the backbone of the nanobrush, with S1 and S2 being complementary, and S3

Oligonucleotide
and S4 being complementary as well.The side arms are depicted in red and blue parts, respectively.

Preparation and characterization of DNA Nanobrush
All DNA nanobrushes were synthesized through a "one-pot" process.Briefly, six oligonucleotides (Table S1) with identical molar concentrations were mixed in 20 mM Tris-HCl buffer (pH 8.0) containing 50 mM MgCl2.The mixtures were heated at 95 o C for 10 min and then incubated on ice for 10 min.The as-prepared nanobrushes were stored at 4 °C for further use.The successful synthesis of the above DNA nanobrushes was validated through polyacrylamide gel electrophoresis (PAGE) gel. 2 μL 6 × Super GelRed Prestain Loading Buffer (US EVERBRIGHT INC.) was added to the above samples and sufficiently mixed.The samples were analyzed by 10% PAGE in 1 × TAE buffer at a constant voltage of 120 V for 50 min.The gel was visualized using a Gel Image System (Amersham Biosciences).
The successful synthesis of nanobrush was also analyzed using atomic force microscopy (AFM).To do this, 1 µM DNA nanobrush samples in a total volume of 50 µL were deposited onto a fresh mica surface.The samples were allowed to adsorb for 15 minutes.After that, 30 μL of ultrapure water was added to the sample and allowed to stay on the surface of the mica sheet for 1 minute.The sample was then absorbed with filter paper, and the washing step was repeated more than 10 times.Finally, the sample was dried with compressed air.The samples were imaged by atomic force microscopy in air scan mode.

Multivalent and monovalent nanobrush binding CEM cells
For monovalent cholesterol-binding CEM cells, 1×10 5 CEM cells were seeded 24 h in advance.After centrifugation to remove the complete medium, 500 μL of 1 × PBS buffer and 100 nM cholesterol-FAM were added and the cells were incubated for 15 minutes at 37 °C.The cells were washed twice with 500 μL of 1 × PBS buffer and then suspended in 1 mL of PBS buffer.To achieve multivalent cholesterol-binding CEM cells, the cholesterol-FAM was simply replaced with nanobrush-FAM.Fluorescence microscopy (Nikon A1R) was used for cell imaging, with the laser excitation wavelength set at 490 nm for FAM.The cell images were observed using a 100× oil objective lens.

Planar nanobrush for two kinds of cell assembly
Human acute lymphoblastic leukemia CCRF-CEM (abbreviated as CEM) cells and human Burkitt lymphoma Ramos cells were cultured in RPMI 1640 medium (GIBCO).The medium was supplemented with 10% fetal bovine serum (FBS) and 1% antibiotics (penicillin-streptomycin-amphotericin B) and the cells were incubated at 37 °C in a CO2 incubator.1×10 6 Ramos cells were mixed in 5 mL of 1640 medium, puted into a culture flask, and cultured overnight for 12 hours.Then, 1 mL of Ramos cells were taken and centrifuged at 1200 rpm for 3 min.The cells were resuspended with 100 μL PBS and stained with CytoTraceTM Red Fluorescent Probe (red).The cells were then adjusted to 1×10 5 for future use.Similarly, for CEM cells, 1×10 5 cells were stained with CellTracker Green CMFDA (green) and used for subsequent experiments.
When the two kinds of cells were assembled, b1-2chol at a final concentration of 500 nM was added to 1×10 5 Ramos cells and incubated for 30 min at 37 °C with a metal bath shaking at 300 rpm.The nanobrush was first anchored to the Ramos cell membrane surface by the hydrophobic effect of cholesterol.Subsequently, 1×10 5 CEM cells were added and incubated for another 30 min at 37 °C with a metal bath shaking at 300 rpm.The nanobrush is again anchored to the CEM cell membrane surface by the hydrophobic effect of cholesterol, and finally, the cell assembly is obtained.
The NIKON A1R confocal microscope observed cell assembly with 490 nm laser and 540 nm laser excitation.The cell images were observed with 100× objective lens.Quantitative data of cell assembly were derived using flow cytometry with 490 nm channel and 540 nm channel excitation.The above operation was repeated 3 times.

Twisted DNA nanobrush for prolonged incubation to build cell clusters
Individual CEM cells to form cell clusters: 1×10 6 CEM cells were mixed in 5 mL of 1640 medium, placed in a culture bottle and cultured overnight for 12 h.Among them, 1 mL of CEM cells was taken and centrifuged at 1200 rpm for 3 min.We resuspended the cell in 200 μL of 1640 medium and adjusted the cells to 1×10 5 for subsequent use.Then CEM cells were added nanobrush (b3-2chol) at a final concentration of 500 nM and placed in the incubator at different times.CEM cells will form stable cell clusters within 24 hours under the combined effect of hydrophobic insertion.
Photographs were taken at different time points using a cell microscope under bright field conditions.We shook the culture flask before imaging to prevent the aggregation of cells during prolonged culture.The magnification of the microscope was 10×.

Working principle of MIET and lifetime-distance conversion
The principle of MIET has been elaborated in our previous publications (ref 1-5).Similar to the fluorescence resonance energy transfer (FRET) process, the gold film can work as an efficient energy acceptor of the excited state energy of a fluorophore.This leads to a strongly distance-dependent modulation of lifetime of the fluorophore over a distance range of ~ 150 nm above the gold surface.This modulation can be calculated by modeling the emitting fluorescent molecule as an electric dipole emitter and solving Maxwell's equation with this source field in the presence of the MIET substrate.Taking into account also the non-radiative transition rate, the observable excited-state fluorescence lifetime (τf) is then found as where (,  0 ) is the emission power of the dipole emitter at the distance of z0 from the substrate surface with orientation angle of , τ0 is the free-space fluorescence lifetime in absence of the gold film, φ represents the quantum yield of the fluorophore, and S0 is the free-space emission power of an ideal electric dipole emitter given  0 =  0 4  2 /3 with c being the speed of light, k0 the wave vector amplitude in vacuum, n the refractive index of water, and p the amplitude of emission dipole moment vector.MIET exploits this lifetime-to distance (τf versus z) dependence for converting measured lifetime values into distance values, see Figure 3 in main text.To calculate this model curve, a priori knowledge of the fluorophore's φ, τ0, and  is required.Previously, we have determined these values for DPPE-atto655: φ = 0.36, τ0 = 2.6 ns.The fluorophore orientation we used for the GUVs measurement is parallel to the membrane and for the cell measurement we used a random orientation.
For the conversion of fluorescence lifetimes into distance values, we used the calculated lifetime-versus-distance MIET curve as described above (see also Figure 3 in main text).For this purpose, a custom-written MATLAB script was used.A MATLAB-based software package for the calculation of MIET lifetime-versus-distance curves as well as the conversion of a lifetime to distance, equipped with a graphical user interface, has been published (ref 6) and is available free of charge at https://projects.gwdg.de/projects/miet.While the published version of the software assumes that the fluorescent emitters are rotating quickly compared to their excited-state lifetime, this was not the case for the measurements in the present work.Here, a dye orientation parallel to the bilayer (and thus to the substrate) was assumed when calculating the MIET calibration curve.

Inter-membrane distance measurement of model membrane
For MIET measurement, 10 μL of 1 μM nanobrushes were added to the SLBs and incubated for 30 min, followed washing by copious buffer to remove the unbonded nanostructures.The diluted GUVs were added onto the nanostructures and incubated for 30 min.

Real-time observation of cell surface engineering process by MIET
Mouse embryonic fibroblast cells (NIH-3T3 cell), human osteosarcoma cells (U2Os cells), and African green monkey kidney cells (COS-7 cells) were cultured in DMEM medium (GIBCO), which was supplemented with 10% fetal bovine serum (FBS) and 1% antibiotics (penicillin-streptomycin-amphotericin B) at 37 °C in a CO2 incubator.
For real-time observation of cells, similar to the GUV measurement, SUVs were deposited on the MIET substrate and fused at room temperature for 30 min to form a uniform lipid bilayer.The substrate was then washed with copious PBS buffer to remove unbound vesicles.Next, 10 μL of 1 μM nanobrush was added to the SLBs and incubated for 30 minutes.After incubation, the substrate was washed with PBS buffer to remove the unbounded nanobrush.Liposome-stained NIH-3T3 cells were then added to the system.Once the cell settled onto the SLB, continuous scanning of the sample was started.A stage top incubator (Cat.No: 12722, Silver Line, ibidi) fits in the microscope, connects to incubator temperature, gas and humidity controllers and creates the proper environment (37 °C, 5% CO2) for live-cell imaging right on the microscope stage.As revealed by confocal microscopy imaging and flow cytometry, the two cells did not assemble without the addition of nanobrush.Similar results also appeared in the samples directly added cholesterol (commercial reagents) or added nanobrush backbone.When ssDNA labeled cholesterol at both ends (2chol-ssDNA) was added, the assembly efficiency was only 5.65%, which might be attributed to the inability of the single strands to form aggregates and fail to bring the two cells closer together.when using single-side arm linked cholesterol (b1-1chol) or double-sided arm linked cholesterol (b1-2chol), the assembly efficiencies can reach 36.8% and 66.4%, respectively.We can adjust the concentration of DNA nanobrush, reaction time and the number of functional groups, thereby controlling the size and proportion of cell assembly.This result shows that even in the presence of 10% FBS, the nanostructures can remain stable in clusters for up to 24 h.

CCRF-CEM cells assembled into clusters with different nanobrush
Figure S7.Aggregates of CCRF-CEM cells were assembled using various nanobrush structures and incubated for 24 hours.(A) The assembled cell groups were characterized using bright-field cell microscopy.Each sample contained the same number of cells.In the case of aggregated cell groups, all cells within the sample chamber tended to aggregate together.Consequently, when focusing on a fixed selection area, the cell density appeared higher due to this clustering effect.(B) Cell aggregation percentage for each nanostructure was quantified after 24 hours of incubation.It's worth mentioning that the assembly ratios (B) are slightly higher than those in Figure S4 due to the extended incubation period.
The results demonstrate the stable assembly of cell clusters by each nanobrush over an extended period.As the DNA backbone underwent gradual twisting, assembly efficiency also progressively increased.We speculate that this twisting exposed the contact sites of each functional chain, leading to more reactive sites, larger contact areas, and higher assembly ratios.The assembly efficiency of the nanobrush with a shortened arm (b4-2chol) was observed to be lower than that of b1-2chol.This disparity is attributed to the reduced length of the side arm, resulting in greater steric hindrance and diminished assembly efficiency.Despite having half the side arm length, the nanobrush (b2-1chol) could still assemble a proportion of cells due to its rotated structure.However, the absence of half the functional chains significantly reduced the assembly efficiency.These findings suggest the capability of distinct nanobrush structures to regulate intercellular assembly spacing, efficiency, and orientation.While the bright-field images provide insights into the nuanced variations in assembly ratios due to different nanobrushes, more comprehensive characterization is essential and should be validated through MIET results.

(Figure S1 .
Figure S1.Schematic diagram of DNA nanobrush.(A) The nanobrush with a backbone 21 bp is a planar structure (b1).(B-D)Changing the backbone bases of the nanobrush to regulate the recognition direction.The nanobrush with a backbone 21 bp is a planar structure.The nanobrush with a backbone 25 bp is a completely twisted conformation.(E) Nanobrush with the short arm where the backbone is 21 bp (b4), both side arm connects cholesterol (b4-2chol).(F) Nanobrush with backbone 22 bp, one side arm connects cholesterol (b2-1chol).

Figure S6 .
Figure S6.Stability experiments of DNA nanostructure-assembled cell clusters at different times.Initially, the experiment was divided into three groups: a blank control; DNA nanostructure backbone (b3, without functional group, 500 nM); b3-2chol (500 nM).As time progressed, cells in the blank and b3 groups dispersed individually withoutforming any noticeable clustered state.In contrast, the b3-2chol group exhibited stable cell assembly, maintaining a clustered cellular morphology.Beyond 24 h, the addition of nanobrush (500 nM) was required to reintroduce stability to the clusters.To facilitate cell growth, RPMI 1640 medium (GIBCO) containing 10% fetal bovine serum (FBS) and 1% antibiotics (penicillin-streptomycin-amphotericin B) was utilized.The cells were incubated at 37 °C in a CO2 incubator.

Characterize the smoothness and
Figure S8.(A) Scanning electron microscopy (SEM) image.(B) Atomic force image (AFM) and (C) the corresponding surface roughness profiles of such MIET substrate.We obtained a root-mean-square value of roughness as 0.8 nm.These surface characterizations affirm that the MIET substrate exhibits exceptional smoothness and uniformity.

Figure S10 .Figure S11 .
Figure S10.Fluorescence intensity images showed the adherence of the GUVs membrane without (A) or with (B) the addition of nanobrush.At least three independent experiments were performed.

Figure S12 .
Figure S12.Cryo-TEM to visualize the inter-membrane distance regulated by DNA nanobrush.(A) Small unilamellar vesicles (SUVs) are depicted without (left) and with (right) the DNA nanobrush (b1-2chol).(B) Quantitative measurement of SUVs distance after adding nanobrush (b1-2chol).Membrane spacing was quantified using Image J, with red curves indicating Gaussian fits of the distance distributions, and the fitted distance is 27.6 nm ± 4.8 nm.

Figure S14 .
Figure S14.Three-dimensional height reconstruction of NIH-3T3 cells at 60 min in the absence of nanobrush.The scale bar is 5 μm.